High-rise window cleaning
Window cleaning of high-rise buildings has long been a dangerous and labour-intensive task, the high risks and costs of this problem demand an inherent safe solution. KITE Robotics, a spin-off of the University of Twente, developed an autonomous window cleaning robot. Currently capable of cleaning medium to medium-high buildings up to 150 m, the Kite Robot covers the demands of the vast majority of (Western-) European market.
Smart inspection
As automated façade cleaning robots replace conventional solutions (gondolas, lifts, suspension bridge systems) an important step for building maintenance is left out: visually inspecting the façade for defects. Defects could mean flaking paint, loose window frames, cracked glass, etc. Currently, facades are inspected by eye from a gondola or even by suspending someone with a rope from the top of the building. The found defects are then manually written down. This is a very time-consuming and often even a dangerous task. For these reasons, the inspection is often skipped, which causes problems that could easily have been prevented. We propose using the KITE Robotics motion system to automatically perform the visual building inspection. Using the robot positioning system and close-up pictures or video of the façade, a complete scan can be stitched together by software. This complete and high-resolution scan can be using to look for defects on the façade. Artificial intelligence could also be applied to the resulting scans to automatically detect defects.
Assignment topics
- Image acquisition, including stabilisation solutions.
- Image processing, combining position and image data to form a scan.
- AI Image processing to automatically find façade defect.
A review of literature, in combination with proof of principle testing, will result in a list of requirements which should then be combined to an integrated design. The assignment is concluded with testing of the finished concept design.
Requirements
- MSc student in Mechanical Engineering, Mechatronics, Electrical Engineering or similar.
- Experienced or interested in deep learning.
- Willing to work in a young company: independence, creativity and taking initiative are highly appreciated.
Contact person
